6 min Devops

Atlassian’s System of Work vision takes shape with Teamwork Collection and Rovo AI Agents

Rovo AI Agents: "one of the best things we've built"

Insight: Agentic AI

Atlassian’s System of Work vision takes shape with Teamwork Collection and Rovo AI Agents

Atlassian Team ’25, the company’s annual event, has a lot of focus on the System of Work framework. This framework means that the various tools used by organizations’ technology teams should be optimally integrated. Teamwork Collection, which Atlassian announced today, fits in perfectly with this. Of course, there is also plenty of room for Rovo AI Agents in this new offering.

System of Work is the framework that Atlassian has arrived at after more than twenty years of experience in the operation of technology teams. The concept mainly revolves around how teams can optimally work together. This means without getting in each other’s way but instead strengthening each other. This does not only involve developers, but also the business side of organizations.

System of Work is not a product that you can purchase from Atlassian. Rather, it’s a way of looking at things like effectiveness, productivity and the overall impact of teams. Naturally, there is a shared basis that is the same for almost all organizations. However, the exact interpretation of System of Work is unique to each company. According to Atlassian, Williams Racing’s (now Atlassian Williams Racing) desire to operate according to the System of Work framework is the reason for becoming the title sponsor of that Formula 1 team. So it’s clear that Atlassian means business.

Something must be done

That something is needed to make organizations faster and more effective seems beyond dispute. Not just from the perspective of Atlassian Williams Racing, which has not done very well in the last twenty years or so. Communication between different teams within many, if not all, organizations that (need to) collaborate on new and existing initiatives in the field of technology leaves a lot to be desired. This is according to research that Atlassian regularly refers to during Team ’25.

Also read: Atlassian Confluence and Jira grow closer, Jira now for business users as well

Atlassian Teamwork Collection

The System of Work framework may sound a bit esoteric, as it is not very concrete. That is why Atlassian wants to flesh it out a bit more today. It is announcing the Teamwork Collection. This collection of apps can be seen as a manifestation of the framework, we hear during a conversation we had with Anu Bharadwaj, President of Atlassian, during Team ’25.

The Teamwork Collection is not really anything new. That is to say, it consists of components that Atlassian already had in its portfolio: Jira, Confluence, Loom and Rovo AI Agents. What is new about the Teamwork Collection is that these components are now better integrated. This is possible thanks to the shared basis of these applications. They all use the same data layer, the Teamwork Graph.

The Teamwork Graph can be seen as the basis of almost everything that Atlassian can offer and does. It is a knowledge graph that Atlassian started about six years ago. It first appeared in Atlassian products about four years ago, we hear from CEO Mike Cannon-Brookes in a session we attended. This knowledge graph not only collects data, but also knows how the objects within an organization – such as teams, messages, goals and the like – relate to each other, according to Atlassian. This makes it a very powerful foundation, especially in combination with Rovo AI Agents.

No Teamwork Collection without Rovo AI Agents

The introduction of Rovo AI Agents, something that Cannon-Brookes calls “one of the best things we have ever built”, has been important for Teamwork Collection and therefore also for System of Work. All data in the Teamwork Graph is now not only available for AI to search or to use as a basis for answers in chats but can also be used to ‘feed’ agents.

The introduction of Rovo AI Agents ensures that a deeper integration between Jira (a system of record), Confluence (a shared workspace) and Loom (a communication platform) actually yields significant results. These are new colleagues, as it were, who can take a lot of work off our hands. Employees no longer have to search desperately for information; agents can do that. They can also take over converting conversations in Loom into diagrams and Confluence whiteboards. In Jira, they can set up entire workflows. Furthermore, they can convert information from meetings into insights that serve as a basis for further actions. And these are only a few of the things Rovo AI Agents can do.

Atlassian has now built more than twenty Rovo AI Agents itself. This includes the agents that make the examples above possible. Organizations can also build AI Agents themselves if they want. Furthermore, Atlassian also realizes that organizations need more than just data from Atlassian environments to make optimal use of AI agents. That is why it has now built more than fifty connectors to third-party applications and environments. To further stimulate the adoption of Rovo AI Agents outside the Atlassian ecosystem, the company has even come up with an additional subscription. For five dollars a month, non-Atlassian users within organizations can also use Rovo AI Agents.

AI has a fundamental impact on Atlassian (and its customers)

All in all, Atlassian is taking a lot of initiative in the field of Rovo AI Agents. It is not waiting for organic adoption within companies, but is showing what is possible with the agents. It is clearly determined to roll out this new part of its portfolio quickly and effectively. The fact that it is including Rovo AI Agents in customers’ subscriptions at no extra cost also speaks volumes. That makes Rovo AI Agents available to all 300,000 organizations that use Atlassian in one fell swoop.

It is clear that Atlassian sees AI and Rovo AI Agents as a crucial component for the success of its System of Work vision. It also helps that Atlassian is used in environments where there are many useful and relatively small-scale applications for AI Agents. This helps enormously in convincing decision-makers within organizations that they should get started with it. This is absolutely necessary if organizations want to (partially) solve the problems they have with effectiveness and efficiency regarding the development of new applications and communication between different teams, both technical and non-technical.

Whether it is all as simple as it seems on paper remains to be seen. The first feedback from customers is reportedly almost unanimously enthusiastic. Atlassian is certainly setting a good example. That is always a good sign, because it gives organizations ready-made options to actually realize a System of Work.